General introduction to stress reconstruction in metal

General introduction to

stress reconstruction in

metal plasticitySam Coppieters & Marco Rossi

The mystery of plastic deformation1

1 E.A. Tekkaya, JSTP International Prize Lecture “The mystery of plastic deformation”,

ICTP 2014, Nagoya, Japan

Stan Hoogenboon, Ankara 2002

Experiment

Step 1: Attach weight until the wire elongates plastically

Step 2: Rotate the weight

Experiment: Step 2

Analysis: Step 1 Uniaxial tension causes yielding and plastic elongation of the wire

F

Material element

Movement of dislocations

Uniaxial Tension:

Phenomenological macroscopic response of metals

Dissipation of Energy

Uniaxial Tension:


Path Dependent

Uniaxial Tension:


Path Dependent

Constitutive relations in incremental form

Analysis: Step 2

F

T

Ingredient 1: Yield criterion

𝜎𝑒 =1

√2𝜎1 − 𝜎2

2 + 𝜎2 − 𝜎32 + 𝜎3 − 𝜎1

212

where are the dislocations?

𝜎𝑒 =3

2𝝈′: 𝝈′

12

Deviatoric stress tensor = Shear = Slip = movement of dislocations

von Mises (1931)

Ingredient 1: Yield criterion

𝜎𝑒 =1

√2𝜎1 − 𝜎2

2 + 𝜎2 − 𝜎32 + 𝜎3 − 𝜎1

212

where are the dislocations?

𝜎𝑒 =3

2𝝈′: 𝝈′

12

Deviatoric stress tensor = Shear = Slip = movement of dislocations

von Mises (1931)

Ingredient 1: Yield Criterion𝜎𝑒 =

1

√2𝜎11 − 𝜎22

2 + 𝜎22 − 𝜎332 + 𝜎33 − 𝜎11

2 + 6(𝜎122 + 𝜎23

2 + 𝜎312 )

12

𝜎𝑒 =1

√22 𝜎𝑧𝑧

2 + 6 𝜏𝑧θ2 )

12

𝜎𝑌 =1

√22 𝜎2 + 6𝜏2)

12

𝜎

𝜎𝑌

2

+ 3𝜏

𝜎𝑌

2

= 1

Taylor, G.O. and H. Quinney. 1931. The plastic distortion of metals.

Phil. Trans. R. Soc., London A230:323

Ingredient 2: Flow Rule• Ingredient 1 = conditions to initiate yielding

• Ingredient 2 = direction of plastic flow

• Normality hypthesis = associated flow 𝑑𝜺𝑝 = 𝑑𝜆𝜕𝑓

𝜕𝝈

0 30 60 90-45

0

45

90

135DIC

p

0 =

0.002

0.02

0.04

0.08

0.115

0.16

0.20

0.24

0.285

von Mises

Hill '48

Yld2000-2d

M=6

M=7

M=7

y

x

D

irection o

f p

lastic s

train

rate

s

/°

Loading direction /°

𝑑𝜺𝑝

Tangent to the yield surface

𝑑𝜀1𝑝

𝑑𝜀2𝑝

Loading point

𝑑𝜺𝑝 =3

2𝑑𝜀𝑒

𝝈′

𝜎𝑒

Consistency condition

von Mises Material

Normality Hypothesis: Validation1

1 Hakoyama et al., Material Modeling of High Strength Steel Sheet Using Multi-axial Tube Expansion Test with Optical

Strain Measurement System, ESAFORM 2013

0 30 60 90-45

0

45

90

135DIC

p

0 =

0.002

0.02

0.04

0.08

0.115

0.16

0.20

0.24

0.285

von Mises

Hill '48

Yld2000-2d

M=6

M=7

M=7

y

x

Dire

ctio

n o

f p

lastic s

tra

in r

ate

s

/°

Loading direction /°

𝑑𝜀1𝑝

𝑑𝜀2𝑝 𝑑𝜺𝑝

Ingredient 3: Strain hardening law

Assume linear strain hardening (Ingredient 3):

→ 𝑑𝜀𝑒𝑝=𝑑𝜎𝑒𝐻

𝑑𝜺𝑝 =3

2

𝑑𝜎𝑒𝐻𝜎𝑒

𝝈′

von Mises (Ingredient 1) and Flow rule (ingredient 2): 𝑑𝜺𝑝 =3

2𝑑𝜀𝑒

𝑝 𝝈′

𝜎𝑒

𝜎𝑒 = 𝜎0 + 𝐻𝜀𝑒𝑝 Slope H

Analysis: step 1

𝑑𝜀𝑧𝑧𝑝𝑙=3

2


𝜎′ =3

2


∙ 𝜎𝑧𝑧 −1

3𝜎𝑧𝑧 =


∙ 𝜎𝑧𝑧 =𝑑𝜎𝑒𝐻

𝜏𝑧𝜃

𝜎𝑧𝑧A B

Initial yield locus

Subsequent yield locus

Analysis: Step 2

(𝑑𝜀𝑧𝑧𝑝𝑙) 𝑠𝑡𝑒𝑝 2=

𝑑𝜎𝑒𝐻 𝜎𝑒 𝑠𝑡𝑒𝑝 2

∙ 𝜎𝑧𝑧

F

T

𝜎𝑒 𝑠𝑡𝑒𝑝 2 =1

√22 𝜎𝑧𝑧

2 + 6(𝜏𝑧θ2 )

12

𝜎𝑒 𝑠𝑡𝑒𝑝 1 = 𝜎𝑧𝑧

→ 𝑑𝜎𝑒 ≈ 𝜎𝑒 𝑠𝑡𝑒𝑝 2 - 𝜎𝑒 𝑠𝑡𝑒𝑝 1> 0

→ (𝑑𝜀𝑧𝑧𝑝𝑙)𝑠𝑡𝑒𝑝 2 > 0

Analysis: Step 2

𝜏𝑧𝜃

𝜎𝑧A B

Initial yield locus

CF

T

→ (𝑑𝜀𝑧𝑧𝑝𝑙)𝑠𝑡𝑒𝑝 2 > 0

Subsequent yield locus

Computational J2 plasticity

von Mises + strain hardening behavior

=> Process-induced modifications/defects

Distribution of defects

𝑑0 = 0.265 𝑚𝑚

𝜀 ≈ 10−41

𝑠

Cu-wire


von Mises + strain hardening behavior: Swift

𝜎𝑒 = 𝐾(𝜀0 + 𝜀𝑒𝑞𝑝𝑙)𝑛

Average behavior


von Mises + strain hardening behavior

𝜎𝑒 = 𝐾(𝜀0 + 𝜀𝑒𝑞𝑝𝑙)𝑛

Extrapolation = uncertainty

FE Analysis: Step 2

FE Analysis: improve understandig

Strong 𝜏 −gradient

FE Analysis: support innovation

Industrial application1

1Becker et al., Fundamentals of the incremental tube forming process, CIRP Annals – Manufacturing Technology 63 (2014) 253-256

Reduction of bending moment

HSS

Effort and predictive accuracy

Effort

Effort Accuracy

Experimental

Computational

Financial

Engineering practice: keep things simple

Plasticity Ductile failure criterion

R. von Mises (1913)

-0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5 Linear

Combined I

Combined II

Combined III

p

p

…but not simpler: Fracture Prediction HET

Hole Expansion Test

High Strength Metal

…but not simpler: Fracture Prediction HET1,2

-0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5 Linear

Combined I

Combined II

Combined III

p

p

Path dependent

RD

1 Benchmark Problem NUMISHEET 20182 Hakoyama, Nakano and Kuwabara, Fracture Prediction of

Hole Expansion Forming Using Forming Limit Stress Criterion, ESAFORM Conference 2017, Dublin.

Plasticity Ductile failure criterion

Experimental Data

Material modeling(Macroscopic model)

Finite element analysis & Fracture prediction

(Macroscopic model)

General Approach

0 200 400 600 8000

200

400

600

800

y

/ M

Pa

x / MPa

0 :1

2 :1

4 :3

1:13: 41: 2

1: 0

1: 4

4 :1

Stress-controlled material testing*

1st quadrant of stress space

=

17 Mechanical tests

*See presentation prof. Kuwabara

Effort and predictive accuracy

Effort

Effort Accuracy

Minimize time-to-market

Reducing traditional testing by leveraging an existing material test Enhance the synergy between experiments and computational methods

MGI1 = effort to discover, manufacture, and deploy advanced materials

twice as fast, at a fraction of the cost.

Strategic Goal: Integrate Experiments & Computation

1Materials Genome Initiative https://www.mgi.gov/

Experimental Data



(Macroscopic model)

General Approach

A B

A. Stress-controlled Material Testing

B. Inverse Material Identification

Multi-Scale Approach

Length scale

Meso ContinuumAtomistic …..Quantum

Stru

ctu

ralIn

teg

rityo

f

me

tal c

om

po

ne

nts

& s

tructu

res

Material Characterization

Material Modelling

Simulation & Validation

Design rules/Standardization

Component testingSelection Strategies

Joining

Intrinsic fracture properties Process-induced fracture properties

Electronic

Structure

Molecular

DynamicsDislocation

DynamicsCrystal

Plasticity

0 200 400 600 8000

200

400

600

800

y

/ M

Pa

x / MPa

0 :1

2 :1

4 :3

1:13: 41: 2

1: 0

1: 4

4 :1

Mircostructurally-informed constitutive modeling1

Virtual Material testing

1st quadrant of stress space

=

17 Virtual tests

=

1 day of experimental effort

1 day of computational effort

Microstructure-based plasticity model(Virtual material test)


Finite element analysis(Macroscopic model) -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6

0

100

200

300

400

500

600

x:y = 4:1

SPCD

Tru

e s

tress

/ M

Pa

Logarithmic plastic strain p

Cruciform

Tubular x

Tubular y

CP

Mircostructurally-informed constitutive modeling1

Microstructure-based plasticity model(Virtual material test)


Finite element analysis(Macroscopic model)

-0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.60

100

200

300

400

500

600

x:y = 4:1

SPCD

Tru

e s

tress

/ M

Pa

Logarithmic plastic strain p

Cruciform

Tubular x

Tubular y

CP

Mircostructurally-informed constitutive modeling

Crystal plasticity simulation(Virtual tensile test)


Finite element analysis(Macroscopic model)

0 200 400 6000

200

400

600 CP(ASR) DH

SPCD

y /

MP

ax / MPa

p

0 =

0.002

0.005

0.01

0.03

0.05

0.10

0.15

0.20

0.25

0.289


Crystal plasticity simulation(Virtual tensile test)



(Macroscopic model)


Experimental Data



(Macroscopic model)

General Approach

A B C

A. Stress-controlled Material Testing

B. Inverse Material Identification

C. Virtual Material Testing

Plastic material response

=

the result of synergistic effects

DC05

Loading conditions

Environmental conditions

Microstructure

…

Synergistic effects during diffuse necking

Large plastic strains,

Biaxial stress state

& strain path change

Strain rate & Temperature increase

Damage evolution

Plastic material response can be complex…

Differential work hardening

Bauschinger effect

Strain rate sensitivity

Kinematic hardening

Ductile Damage

…

Plastic material response can be complex…


Bauschinger effect

Strain rate sensitivity

Kinematic hardening

Ductile Damage

…

Isotropic work hardening

𝜎

𝜀


Experimental Data


Finite element analysis (Macroscopic model)

Problem statement

Problems

Many models exist

Anisotropic Yield Functions

Hill 1948

Hill 1990

Gotoh

Barlat Yld2000-2d

Barlat Yld2004

Yoshida

BBC2005

BBC2008

Vegter Spline

Karafills-Boyce

CPB2006

…

Coupled Damage Models

Experimental Data



Problem statement

Selection strategy

Identification strategy

Problems

Many models

Need

Experimental Data



Problem statement

(Selection strategy)


Problems

Many models

Need

Models not available

in commercial FE code Implementation

Experimental Data



Problem statement



Problems

Many models

Need



Implementation = stress reconstruction

FE CODE

Stress integration – Stress reconstruction -Stress update

Stress update algorithm

Experimental Data



Problem statement



Problems

Many models

Need



Many codes

Unified Material Model Driver for plasticity1

1JANCAE, The Japan Association for Nonlinear CAE

Experimental Data



Problem statement



Problems

Many models

Need



Many codes

Inverse Material Identification

Numerically converged value (e.g. FEMU)

Experimentally measured value (e.g. VFM)

Finite Element Model Updating (FEMU)

Finite Element Model Updating (FEMU)

Virtual Fields Method

Global Equilibrium

Energy Method

V (DIC)

y

xx

y

Energy MethodDIC

(U,V)STRAIN RECONSTRUCTION STRESS RECONSTRUCTION

LE22 S22

∆𝜀𝑖

Converged steps in FEA

During a quasi-static test the internal work equals the external work

Lines A-B and C-D remain straight The Volume V is constant The stresses and strains are constant

through the thickness of the sheet

y

x

𝝈𝒊𝒋 , 𝜺𝒊𝒋 V, S, t

A B

C D

𝑓𝑖 , 𝑢𝑖

Energy Method

Forming Limit Curve

Forming Limit Curve*

* Junying Min et al., Compensation for process-dependent effects in the determination of localized necking limits,

International Journal of Mechanical Sciences117(2016)115–134

Compensate for nonlinear strain path

Documents

General introduction to stress reconstruction in metal